r 

The Epistemological Theory of Argument-
How and Why? 

CHRISTOPH LUMER University a/Siena, Italy 

Abstract: The article outlines a general 
epistemological theory of argument: a theory 
that regards providingjustified belief as the 
principal aim of argumentation, and defends 
it instrumentalistically. After introducing 
some central terms of such a theory (2), 
answers to its central questions arc 
proposed: the primary object and structure 
ofthe theory (3), the function of arguments, 
which is to lead to justified belief (4), the 
way such arguments function, which is to 
guide the addressee's cognizing (5), objective 
versus subjective aspects of argumentation 
(6), designing different types of argument 
(7). Then the notion of '(argumentatively) 
valid argument' is defined and criteria tor the 
adequate use of such arguments are 
introduced (8). Finally, this conception is 
justified as, among others, leading to more 
true beliefs than competing conceptions (9). 

Resume: On expose les grandes lignes 
d'une theorie epistemologique generale 
des arguments: une theorie selon laquelle 
Ie but principal de I'argumentation est 
d'avancer des croyances justifiees, et on 
soutient celte theorie tonctionnellement. 
Apres I'introduction de termes importants 
de celte theorie (2) on discute des themes 
qui repondent aux questions importantes: 
I'objectif principal et la structure d'une 
theorie (3), la tonction des arguments, qui 
est de mener a des croyances justitiees 
(4), la fayon que les arguments fonction-
nent, qui est de guider la retlexion de 
I'auditoire des arguments (5), Ics aspects 
objectifs et subjectifs de I'argumentation 
(6), la construction de differents types 
d'arguments (7). Ensuite on definit la 
notion de validite argumentative, et on 
introduit des criteres pour I'emploi 
adequat de tels arguments (8). Entin, on 
justifie celte conception, parmi d' autres, 
en demontrant qu'elle mene a un plus 
grand nombre de croyances vraies que Ics 
conceptions opposees (9). 

Keywords: argument, argumentation, epistemological approach, Practical Theory of 

Argument,justified belief, acceptability, epistemic accessibility, guiding cognition, 
objectivity, validity, adequacy, justification, function of argument 

1. Defining 'Epistemological Theory of Argument' and the Structure 
of this Paper 

An epistemological theory of argument is characterized by two features. I. It 
takes the standard function of arguments to be: to lead the argument's addressee 
to (rationally) justified belief, i.e" to guide him to realize the truth or acceptability 
of the argument's thesis-where 'acceptability' is intended to be a broader term, 

©Informal Logic Vol. 25, No.3 (2005): pp. 213-243. 



214 Christoph Lumer 

meaning: truth, high probability or verisimilitude. 2. It develops criteria for good 
arguments and argumentation on this basis, i.e., it designs them in such a way as 
to fulfil their epistemic function. This is not the broadest characterization of an 
epistemological theory of argument but it still leaves room for quite a wide range 
of specifications. In the following I will be more restrictive and introduce more 
specific determinations, which will be justified in the subsequent sections. I use 
the label "Practic!ll Theory of Argument" to distinguish this more specific theory. 1 

The paper is structured as follows. In section 2 some central notions of the 
epistemological theory of argument are explained just in order to clarify their intended 
content, without justificatory aims: 'argument', 'argumentation', 'justified belief', 
'cognition', 'epistemology', 'standard function', etc. In sections 3-7 some central 
inner-theoretical questions of an epistemological theory of argument are discussed 
and the respective determinations of the Practical Theory of Argument are introduced 
and defended: the primary object (argument, argumentation or discussion) and the 
structure of the theory as well as the ontology of arguments (sect. 3), the exact 
function of arguments (sect. 4), their way of functioning (sect. 5), the problem of 
objectivity versus subjectivity of argumentation (sect. 6) and the basis for different 
types of arguments (sect. 7). After these discussions the concepts of' argument', 
'(argumentatively) valid argument' and of 'situational adequacy' can be defined 
(sect. 8). Having thus completed a rather detailed picture of the Practical Theory 
of Argument, finally, this theory and the epistemological approach to argumentation 
in general are justified in comparison to competing approaches: the rhetorical and 
the consensus theory of argumentation (sect. 9). 

2. Some Central Notions of the Epistemological Theory of Argument 

An argument here is intended to be a sequence of statements or judgements (i.e., 
propositions plus the assertive mode 2), one of which is the thesis or conclusion, 
while the others are the r!"asons or premises, and of an indicator of argument such 
as 'therefore', indicating which proposition is the thesis and what are the reasons. 
Statements "say" that the enclosed proposition is true. Only propositions can be 
true or false, and e.g. logical implications hold only between propositions. But 
sometimes, in a loose way of speaking, I will also say that a "statement / thesis is 
true" etc. instead of saying the "statement's / thesis's proposition is true". 
"Argumentation" here shall mean: the (mostly monologcial) act of arguing, i.e., the 
presentation of an argument. In addition there are discussions, i.e., dialogues 
consisting mainly of argumentations by the participants. Epistemological approaches 
to argumentation take arguments to conceptually and systematically precede 
argumentation and discussion, which does not exclude that they also study the use 
of arguments in discussions (see, e.g., Lumer 1988). Their central task is to establish 
precise criteria for good arguments and their use. (Other tasks of an epistemological 
theory of argument are: 1. "self-reflection", i.e., fixing the aim and methods of the 
theory of argument, defining the basic concepts of the theory, in particular 



The Epistemological Theory of Argument-How and Why? 215 

'argument', determining the function of arguments and explaining their way of 
functioning. etc.; 2. to develop a theory of interpretation of argumentative utterances; 
3. to develop a theory of fallacies and 4. a theory of argumentative discussions, 
etc. (see Lumer 2000b, sects. 3-7).) 

The other central notion of the theory, 'justified belief, includes knowledge, 
which roughly may be identified with (surely) justified true belief, but also justified 
beliefs that have been acquired by rational but uncertain means and therefore may 
be false. Justified beliefs that are not knowledge aim at truth, too. But, because of 
the limited means they are forced to rely on, they do not always achieve this goal. 
Justified belief is the result of (rational) cognition or cognizing, i.e., a process of 
acquiring beliefthat satisfies certain epistemological standards; and these standards 
require checking as to whether certain conditions of the truth or acceptability. of 
the proposition in question are fulfilled. Justified belief then consists of two 
components: the belief resulting from a positive outcome of this check and the 
subjective justification, i.e., the memory of at least some traces of the cognition. 
The subjective justification is rationally necessary because there are false cognitions 
and because we use uncertain ways of adopting beliefs. As a consequence, this 
sometimes results in conflicting, inconsistent beliefs, and we have to revise at least 
one of them, namely the most weakly justified belief. And it comes to situations 
where a belief just revised had served for justifying another belief, so that this 
second belief now has to be revised as well. In such situations, now disposing of 
subjective justifications enables the person to assess the strength of the evidence 
had for adopting the belief (in case of conflicting beliefs), to then relinquish the 
weaker justified belief, and to identify beliefs that depend on a revised belief. 

"Epistemology" here means a normative theory that among others provides 
criteria for rational justification, i.e., it provides the standards to be followed in 
rational cognition. (This is a rather broad concept of 'epistemology', which includes, 
among others, also logic and probability theory.) These standards have to be 
developed on the basis of truth definitions for certain types of propositions. The 
deductive epistemological principle 'a propositionp is true ifp is logically implied 
by true premises', e.g., in the end relies on the truth definitions for complex 
propositions (i.e., propositions containing logical operators): the "mechanics" of 
the logical operators guarantees what is expressed in the deductive epistemological 
principle. Justifying epistemological principles on the basis of truth definitions is 
important because only this guarantees that fulfilling the criteria given by an 
epistemological principle for the truth or acceptability of a proposition really implies 
the truth or (at least) acceptability ofthat proposition. "Acceptability" of a proposition 
here always is intended to be epistemically rational acceptability, i.e., truth, high 
probability or verisimilitude. 

Finally, when speaking of the "standard function" of arguments, "function" is 
meant in the system-theoretical sense: relation between input and output. There is 
a structure, i.e., the expression of an argument, that if "fed" with some input, such 



216 Christoph Lumer 

as being presented to an addressee who does not (sufficiently) know the thesis to 
be true, under certain conditions produces a certain output, e.g., the addressee's 
justified belief. Arguments have many functions, they can impress, bore, amuse 
people, etc. The standard function of arguments shall be the function for which 
good arguments have been designed in the theory of argument, namely to lead to 
justified belief. 

3. The Primary Object, Scope and Structure of the Theory 

Above, 'argument', 'reason', 'argumentation' and 'discussion' have been 
distinguished. Some theories of argumentation have argumentation or discussion 
or reasons as their primary objects of study. Of course, there are arguments, 
argumentations as well as discussions and we need concepts, expressions and 
theories for all of them. However, as already stated, epistemological approaches to 
argumentation primarily deal with arguments and their use; and this is necessarily 
so. Or more precisely: 1. 'Argument' is taken as the systematically primary concept 
in terms of which the concepts '(argumentative) reason' and 'argumentation' are 
defined; subsequently' argumentative discussion' is defined with the latter terms. 
2. In addition, the epistemological theory of argument deals with the respective 
objects in the same systematic order, which-apart from 'reason' which is only 
an auxiliary term-reflects their increasing complexity: argument, argumentation, 
argumentative discussion. (This does not imply that' argument' is taken as a primitive 
notion. Of course, it shall and will be defined-but in terms of concepts like 
'statement', which are not specific to argumentation theory.) The various parts of 
this claim shall now be justified. 

'Argument' systematically precedes' reason'. At a first glance just the contrary 
may seem to be true because arguments are mereologically composed of a thesis, 
an indicator of argumem and reasons for the thesis, so that the straightforward 
strategy to define 'argument' seems to be: an argument is a statement plus reasons 
for it .... But, contrary to this first appearance, the term '(good) reason' can 
sensibly be defined only by using the tenn 'argument'. First, argumentative reasons 
are always reasons for a thesis; a one-digit concept of reason would make every 
statement a reason, so it would be a useless concept. Thus, the thesis has always 
to be added to the reason. Second, whether a reason (in the sense of a single 
statement) is a good reason for a given thesis depends substantially on the other 
reasons of the argument, so, in the end, on the whole argument. This is so because 
if, e.g., a reason p does not in itself imply the thesis q, it is not clear in what relation 
p could stand to q for making it a reason for q. So the other reasons have to be 
taken into consideration too; and the best way to define '(argumentative) reason' 
is to equate them with the statements of an argument that are different from the 
thesis. 

'Argument' and 'argumentation' systematically precede also 'discussion' or 
'argumentative discourse'. The kernel of rational cognition, i.e., acquisition of 

, 



r The Epistemological Theory a/Argument-How and Why? 217 
justified belief, is essentially an internal. mental process: the subject examines 
whether criteria for the acceptability of a proposition are fulfilled. And ideally, the 
reasons' part of an argument consists of statements asserting that these criteria 
are fulfilled. Of course, cognition can be furthered or even hindered by other 
people's comments, hints, arguments, etc., i.e., (pieces of) dialogue. However, 
proper cognition takes place internally and consists in checking the fulfilment of 
criteria for the acceptability of a thesis; other people's talk here has only the role of 
prompting (or preventing) certain check-ups, it does not make up these check-
ups. In principle the whole cognitive process can be-and often really is--executed 
independently of other people's intervention. (Observing, preparing observations, 
conducting experiments, getting information from other people about their 
experiences and reflections go far beyond mental processes. But to acquire the 
respective justified belief all this has to be internally represented, and the proper 
decision about what to do with this information, if and how to transform it into 
what kind of justified belief, is a mental process.) All this means that a linguistic 
device that mirrors cognition and its epistemic result is ""mono logic", namely some 
sort of sequence of assertions that sufficient criteria for the acceptability of the 
thesis are fulfilled. And, because cognition is at the centre of the epistemological 
approach, this excludes discussions as its primary object of study. In a second 
step though, arguments and argumentations can be integrated in discussions. The 
kernel of (rational) discussion is argumentation, i.e., the presentation of an argument. 
But discussions are then enlarged by questions, criticism, counter-claims, retractions, 
etc. (see Lumer 1988, 452-461). 

The most difficult part of the present justification is to shO\", the conceptual 
and systematic precedence of arguments over argumcntation. Argumentations in a 
certain sense are more fine-grained than arguments. This can be seen from the 
fact that a complete description of an argumentation has to include some description 
of the argument but it also has to specify some further variables (for the situation. 
the expression and perhaps for the addressee and the purpose): person s at time ( 
expressed the argument 'a' in language 1/ uttered "u" (which has the meaning '(/) 
for convincing h. The main reasons for taking arguments as the central notion and 
primary object of study in an epistemological approach to argumentation are these: 

I. 'Argument' is the more primitive notion. One can easily define' argumentatio'n' 
in terms of 'argument' and vice versa. (An argumentation is a complex of 
speech acts in which an argument is presented. Versus: An argument is the 
meaning of an actual or possible argumentation.) But if one then tries to define 
'argument' without making use of the notion of' argumentation' or j'ice j'crsa it 
turns out that such a definition of 'argumentation' must, in an oblique \Nay, use 
the defining conditions of 'argument', whereas such a definition of' argument' 
need not recur to the defining conditions of an argumentation. (The definition 
of '(deductive) argumentation' would have to go along these lines: " ... sequence 
of speech acts uttered by someone; ... ; one of the speech acts expresses the 



218 Christoph Lumer 

thesis, i.e., an assertion whose proposition the arguer wants to show to be true; 
the other speech acts express assertions whose propositions are true and logically 
imply the truth of the thesis's proposition." The definition of 'argument' would 
go along these lines: "sequence of assertions; ... ; one is the thesis; the other 
assertions' propositions are true and logically imply the thesis's proposition." 
So trying to use 'argumentation' as the more primitive notion is awkward at the 
least.) 

2. There are uses of arguments independent of argumentation. The typical or 
standard use of arguments is their presentation in an argumentation mostly for 
convincing. However, this is not their only use (see below, sect. 4); and some 
of their uses are independent of argumentation but not vice versa (because of 
no. 1). 

2.1. Arguments are ideal forms of subjective justifications. They list the conditions 
that make the thesis acceptable and have been checked-with a positive result--
by the subject.The additional information that is implied in describing 
argumentation-who uttered the argument, at what time-, is irrelevant for 
subjective justification. Cognition is limited by situational constraints; in particular 
the epistemic subject must have cognized J the reasons as acceptable and he 
must understand the argumentative relation. But, first, the original arguer and 
the time of argumentation are not parts of these situational constraints. Second, 
it is not even necessary that there is such an arguer; the epistemic subject may 
have found the argument by himself. Third, statements that situational constraints 
are fulfilled are not parts of the subjective justification. Such constraints shall 
mainly guarantee that the cognition iri question is accessible to the epistemic 
subject. Ifhe has cognized the thesis along the line of the argument, the respective 
constraints obviously were fulfilled; he must not additionally check that they 
were fulfilled, and there is no need to include the statements about their fulfilment 
in the subjective justification. 

2.2. Another use of arguments without argumentation is inquiry (cf. Meiland 1989, 
194). An epistemic subject tries to develop a valid and sound argument that 
could prove a certain conjecture. In such cases there is no arguer who presents 
the argument to somebody else. And even in a very broad sense the subject 
does not present the argument to himself; he constructs a hypothetical argument 
instead and examines its validity and soundness. 
The general gist ofthese arguments in favour of the claim presented above that 

arguments conceptually and systematically precede reasons, argumentation and 
discussion is this. An epistemological approach to argumentation constructs 
arguments in view of their epistemic function as the statements that formulate 
what has to be (and later has been) checked for cognizing the thesis. This epistemic 
kernel in principle does not depend on argumentation or discussion. But, the other 
way round, this epistemic kernel is also the kernel of what happens in successful 
argumentation and discussion. 



The Epistemological Theory of Argument-How and Why? 219 

4. Instrumentality and the Function of Arguments 

Arguments are abstract entities, which exist independently of a subject who, e.g., 
has designed them or brought them into being. Therefore, arguments as such 
cannot have aims or have been designed as means. There are many arguments that 
will never be thought by anybody, and there are many arguments expressed by 
several people with perhaps quite different aims. (Argumentations, being acts, of 
course can have aims; they are produced by a subject for a certain purpose, and 
their aim resides in the subject (see Johnson 2000, 168).) Nevertheless, arguments 
as such are functional entities, instruments that shall be useful for some aim. The 
way here to conceptualize this is to say that they have a standard (or proper) 
function. "To have a function" does not mean that the function must have been 
performed; it is sufficient that in case the inputs were fed in the respective structure, 
the appertaining output would be produced. Arguments have a lot of functions, 
i.e., input-output relations. But the criteria for good arguments in the epistemological 
theory of argument have been designed in such a way as to fulfil a very specific 
function, namely the standard function (or more precisely: they are designed to 
fulfil several closely related functions) (Lumer 1991, 99f.). 

The standard input of arguments is that one of their linguistic expressions is 
presented to a linguistically proficient, open-minded, attentive and discriminating 
addressee who does not yet have a sufficiently justified belief in the thesis. The 
standard output of arguments is that the addressee justifiedly believes that the 
thesis is true or acceptable (or that he has a better justified belief in the thesis). The 
standard function, i.e., the relation between standard input and standard output, 
may be termed "leading to justified belief' or to "rationally convince" (Lumer 
1990, 43 f; 1991, 1"00). 

Some further functions, closely related to this standard function and still specific 
to arguments, are these (see Lumer 1990, 49f; see now also Blair 2004, 139-141). 
1. Arguments can represent justified beliefs in the sense of being the ideally formed 
content of justified belief. In section 2 I said that justified belief is acquired by 
cognition and that it consists in believing in the thesis and remembering at least 
traces of the cognition, and at best remembering the results of the single check-
ups of the thesis's conditions of acceptability. In case of inferred beliefs, the 
content of a justified belief (thesis and ideal justification) is identical to an argument. 
The input of the function 'to represent justified belief', then, is to cognize the 
thesis, and its output is to believe in the argument-which is a certain form of 
justified belief 2. An arguer may combine the standard function and the function 
just explained by convincing an addressee with the help of an argument that makes 
up the content of her own justified belief. This function may be called "transfer of 
justified belief'. What is important about transfer of justified belief is that the 
arguer does not only communicate what she believes or induce the addressee to 
believe the same, but that she transfers justified belief as such; the addressee 
acquires the same justified belief that the arguer already has. 3. Arguments can be 



220 Christoph Lumer 

used for autonomous (argumentative) cognition. A person tentatively constructs 
an argument for some hypothesis she does not yet justifiedly believe (to a sufficient 
degree) to be acceptable and then examines whether this argument is a good one. 
If it turns out to be a good one, she has already cognized the thesis's acceptability. 
(Why this is so will be explained in sect. 6.)44. Reconsidering one s justified belief 
for securing it presupposes that a person already has some justified belief of the 
thesis or at least believes in the thesis. And she then uses an argument found 
soil1ewhere to recheck criteria for the thesis's acceptability. In case of a positive 
outcome she has sec ured her j usti fied be 1 ief, maki ng it more certai n. 5. 
Intersuhjective securing 0/ a justified belie/begins with disclosing one's justified 
belief by presenting the respective argument to some audience for intersubjective 
critique. In this way the private, subjective justification is made public and thus 
can be scrutinized by others who may reveal errors or other insufficiencies. If 
one's justified belief has passed this test one can be more certain about it (see 
Lumer 1988, 448-450). 

An instrument that fulfi Is its standard function is called '"functioning" (or '"in 
working order"). As a more specific expression for the functioning of arguments, 
it will here be said that the argument is '"(argumentatively) valid'. This is not the 
same as logically valid. In cases of deductive arguments, argumentative validity 
includes logical validity but goes beyond it; in cases of non-deductive arguments, 
argumentative validity does not even include logical validity. The conditions that 
have to be fulfilled for argumentative validity will be specified below. 

Functioning instruments can be used in various situations. To fulfil their standard 
function they have to be used in an adequate manner; in particular the standard 
input has to be provided. A functioning instrument that is not used in an adequate 
way may be as good as possible but still not provide the desired output. If you drill 
into concrete using a bit made for drilling into wood, you will not produce the 
desired result (but ruin your drill bit). Instructions for use should describe the 
appropriate manner for using the instrument to fulfil its standard function. The 
same holds for good, (argumentatively) valid arguments. In order to fulfil their 
standard function, i.e., for rationally convincing, they have to be used in an adequate 
manner. I fthe addressee, e.g., does not justifiedly believe some premise to be true, 
he will not be convinced by the argument--even if the argument is (argumentatively) 
valid, e.g., sound (deductively valid with true premises). Therefore the conditions 
for argumentative validity (i.e., for good arguments) have to be supplemented by 
conditions for an adequate use of arguments in certain situations (conditions for 
good usc). So there will be two sets of conditions for good argumentation, conditions 
for good (argumentatively valid) arguments and conditions for their good 
(situationally adequate) use. The latter conditions are instructions for using 
arguments. 



The Epistemological Theory of Argument-How and Why? 221 

5. Arguments' Way of Functioning 

In the foregoing section the standard function of arguments (leading to justified 
belief, rationally convincing) has been explained, and their standard input and output 
have been specified. What still has to be explained, and what is very important 
from an epistemological point of view, is their way of functioning: How do arguments 
fed with the standard input produce the standard output? How does rational 
convincing work? 

The way arguments function can be explained most easily for deductive 
arguments (see Lumer 1990,45-48; 280-281; 1991, 102-104). To fulfil their standard 
function arguments have to be argumentatively valid and they must be used 
adequately. The following are abridged versions of the two respective sets of 
criteria. 5 

Argumentative validity of deductive arguments: 

DAD: Domain: The argument consists of I. a single statement, the thesis, 2. an 
indicator of argument (like 'therefore', 'for this reason'), and 3. a set of further 
statements, the premises. 

DA1.· Indicator: The indicator indicates I. that the whole sequence is an argument, 
2. which statement is the thesis, and 3. which statements are the premises. 

DA2: Guarantee of truth: I. The premises' propositions are true, 2. and they 
logically imply the thesis's proposition. 

DA3: Adequacy in principle: There is at least one person who justifiedly believes 
that the premises are acceptable but who does not justifiedly believe this about the 
thesis. 

Situational adequacy of deductive arguments: 

DA4: Situational adequacy: An (argumentatively) valid deductive argument "is 
adequate for rationally convincing an addressee of the thesis if the following 
conditions are fulfilled. I. The addressee is linguistically proficient, open-minded, 
attentive, and discriminating. 2. He knows that the premises are true but does not 
know if the thesis is likewise. 6 3. The relation of implication between the premises' 
propositions and the proposition of the thesis is sufficiently direct so that it can 
easily be grasped by the addressee. 

How do adequately used valid arguments (defined by the above conditions) 
lead to justified beliefs? The most concise answer is: they guide the addressee in 
his process of cognizing the thesis. More comprehensively and precisely what 
happens is this: The indicator of argument informs the addressee that this is an 
argument and what its thesis is. In addition the indicator calls his attention to the 
fact that with the help of this argument he-at least in the arguer's opinion--can 
cognize the thesis's acceptability. The addressee now may accept this invitation 
and begin to examine the thesis and the premises. 

If the argument is clear, the systematically first step of this examination is to 
find out the epistemological principle upon which the argument is based. All 



222 Christoph Lumer 

(argumentatively) valid arguments rely on an efficient epistemological principle, 
such as the deductive epistemological principle: 'a proposition is true ifit is logically 
implied by true premises' or the genesis o/justified belie/principle: 'a proposition 
is true if it has been correctly verified'. Such epistemological principles are general 
criteria for the truth or acceptability of a wide range of propositions. An 
epistemological principle is efficient iffulfilment of its conditions really guarantees 
the truth or acceptability of the respective proposition, i.e., ifthe principle (mostly) 
leads to true or truth-like conclusions. The revelation principle, 'a proposition pis 
true ifit is stated in the Bible', e.g., is not efficient. The efficiency of epistemological 
principles relies on their relation to the truth definitions for the propositions in 
question. In the case of the genesis of justified belief principle, this relation is 
established by verification. "To verify" just means to check whether the truth 
conditions of some proposition are fulfilled and to find that they are. Different 
types of arguments have to be distinguished according to the epistemological principle 
they rely on: deductive arguments, genesis of justified belief arguments, (various 
kinds of) probabilistic arguments, practical arguments, etc. (see below, sect. 7). 
Most people are not able to formulate epistemological principles. Nevertheless, if 
trained a bit they have an intuitive understanding of such principles in the sense 
that they know which conditions have to be fulfilled by a given type of argument. 7 
To find out which epistemological principle the argument relies on is the key for 
using the argument as a guide to cognition. Without recognizing (at least in the 
weak, intuitive sense just explained) the underlying epistemological principle, one 
cannot make sense of the argument; one does not know which kind of standards 
the argument is supposed to show to be fulfilled. The indicator of argument 
sometimes also indicates the respective epistemological principle and the type of 
argument; "from this follows" may be such an indicator, though it is not unequivocal, 
indicating deductive inferences as well as deductive arguments. Unfortunately, 
this is a rare case and usually the addressee has to find out the epistemological 
principle from semantic CL,,~S of the thesis and the reasons. In complete deductive 
arguments, e.g., the thesis's notions already appear in the reasons; in complete 
practical arguments, the thesis is a value judgement and reasons consist of 
statements about the value object's implications and their respective values. 8 

Once it is found, the underlying epistemological principle can be used as a 
checklist for cognizing the truth or acceptability of the thesis. The addressee now 
has to check whether all the conditions for the truth or acceptability of the thesis 
listed in this principle are fulfilled. In the case of deductive arguments, this means 
checking 1. whether certain premises are true and 2. whether they logically imply 
the thesis. But which premises? The premises specified in the argument's reasons, 
of course. If the argument is argumentatively valid and adequate, the addressee 
can immediately check whether those two conditions are fulfilled, and the result of 
this check-up will be positive. 1. The premises are true (DA2.1), and the addressee 
knows them to be true (04.2). So when confronted with the premises, the addressee 
will in each case immediately recognize them as true. 2. The premises logically 



The Epistemological Theory of Argument-Ho'wand Why? 223 

imply the thesis (DA2.2), the addressee is linguistically proficient (DA4.1), which 
includes linguistic proficiency with respect to the logical operators and thus also 
logical implications, and the relation of implication in the argument is sufficiently 
direct so that he can easily grasp them (DA4.3). All this enables the addressee to 
immediately check that the premises imply the thesis and to come up with a positive 
result. 

The last step of the cognition, then, is rather simple. Because the addressee 
knows about the epistemological principle, he can infer that the two conditions for 
the thesis's acceptability are fulfilled so that the thesis must be acceptable. 

It has been said that (argumentatively) valid and adequate arguments guide the 
addressee in his process of cognizing the thesis. This guidance consists of several 
things: specifying the thesis that will be examined, i.e., informing that this thesis 
can be proved; specifying the epistemological principle that may be used for 
cognizing the thesis's acceptability; and specifying the premises from which the 
thesis can be inferred (the epistemological principle in itself, being completely 
general, does not yet do this). By specifying all these things, the argument invites 
the addressee to examine what has to be examined for cognizing the thesis's 
acceptability. And, in an (argumentatively) valid and adequate argument, the 
specifications are chosen in such a way that what has to be examined can immediately 
be checked for a positive result; this is the point of good argumentation. Formulated 
in an overly subtle and somewhat exaggerated way, guidance of cognizing by 
good argumentation works like this: 

"Do you want to cognize the acceptability of thesis tT' 

"Yes." 

"Okay. The cognition I have to offer is based on the deductive 
epistemological principle. So please take this as your checklist! The first 
condition on this list should be that the premises are true. So here are the 
premises: PI' P

2
, ... , PlIo Please check whether they are true!" 

(Because these premises are chosen in the right way the addressee does 
so with a positive result.) . 

''The second condition on your checklist should be logical implication. 
So please check whether PI' P

2
' ... , P" imply t!" 

(Because the implication is sufficiently direct the addressee does so with 
a positive result.) 

"Now you have ascertained that all the conditions of your checklist for 
the acceptability of t are fulfilled. So now please accept t!" 

Of course, argumentation (usually) does not take this explicit and slightly exaggerated 
form. But the point of my explanation of the way it functions is that the usual good 
argument contains all the information needed to inflate it to this form. 

Another crucial point of this conception of arguments' way of functioning is 
this. Even ifthe addressee is guided in cognizing the acceptability of the thesis, he 



224 Christoph Lumer 

himself takes the essential steps in the cognitive process. He uses the epistemological 
principle as a checklist; and he checks wheether its conditions are fulfilled. So he 
convinces himselJofthe thesis's acceptability. And the epistemological approach 
to argumentation explains the rationality of rationally convincing someone by 
argumentation this way: the arguer convinces the addressee by guiding him in a 
process of convincing himself by using epistemological standards. 

Truth dejinitio,?s imply rather specific ways of (directly) verifying the defined 
proposition. For directly verifying, e.g., a proposition 'p&q' one has to verify that 
p and to verify that q. Epistemological principles-though being based on truth 
definitions---often are much less specific and allow many (perhaps infinite) indirect 
forms of cognition for one proposition. The deductive epistemological principle, 
for example, contains an existential quantification: 'If there is a set of true 
propositions PI' ... ,p" that imply a proposition t then t is true.' If t is true, there is 
an infinity of such sets. For t = p&q, {(P&q) V r, ~r} may be such a set. But most 
of them are not epistemically accessible, i.e., we do not justifiedly believe their 
elements to be true (if we do not already justifiedly believe t to be true). So what a 
good (in this case deductive) argumentation does is to choose one of these sets 
that is epistemically accessible to the addressee. An epistemological approach to 
deductive argumentation is sometimes criticized because deductive argumentation 
allegedly cannot provide new knowledge because all the information to be learned 
is yet implied in the premises, which have already to be accepted. Of course, the 
thesis is implied by the premises; the latter are true---otherwise a deductive argument 
would not be (argumentatively) valid; and the addressee has already accepted the 
premises---otherwise using this argument would not be adequate for convincing. 
Nonetheless, adequate deductive argumentation provides new knowledge about 
the thesis. Before the argumentation, the addressee did not justifiedly believe the 
thesis to be acceptable (see DA4.2), but afterwards he does. The reason for this is 
that we cannot know all the implications of the justified beliefs already acquired 
(our justified beliefs are nL't deductively closed). First, there is an infinity of such 
implications, so that for a finite brain it is physically impossible to know them all; 
second, most of them are irrelevant. Adequate (deductive) argumentation in this 
situation is informative in a secondary sense: it tells the addressee that by starting 
from this set of premises he can cognize the acceptability of the thesis. This 
secondary information saves the addressee the work of finding an adequate set of 
premises on his own from which the thesis can be inferred. This secondary 
information does the above-mentioned work of transferring justified belief as such 
from the arguer to the addressee. 



r 
The Epistemological Theory of Argument-How and Why? 225 

6. Argumentative Validity and Adequacy-a Synthesis of the 
Objectivist and Subjectivist Accounts of Epistemologically Good 
Argumentation 

One of the major problems an epistemological approach to argumentation has to 
resolve is the relation between the objectivity and subjectivity of argumentation. 
On the one hand, the epistemological approach aims at objective knowledge and 
true beliefs: good argumentation has to be designed in such a way that if everything 
goes smoothly the result is that the addressee knows a true proposition to be true 
or justifiedly believes an acceptable proposition to be acceptable. Good arguments 
have to guarantee the truth or acceptability of the thesis. And this distinguishes the 
epistemological approach from a rhetorical approach: an argument can be good in 
this specific sense of the epistemological approach even if it did not convince (for 
instance, because the addressee was stubborn or not sufficiently attentive or did 
not believe some premise), and it can be bad in this sense even though it did 
convince because it was persuasive (Feldman 1994, 168; Lumer 1990, 29). On 
the other hand, according to the epistemological approach, argumentation will lead 
to justified belief, guide cognition, etc., which is an epistemic task. But cognition is 
an activity of a subject and depends on the subject's epistemic situation. There are 
several subjective aspects in good argumentation. 

1. Epistemic accessibility of reasons and inferences: First, the premises of an 
argument can be true but the addressee does not justifiedly believe this, or the 
inference may be too difficult for this addressee; then this is not a good argument 
for convincing this addressee. But it could be a good argument for convincing 
another addressee who justifiedly believes the premises to be true and, e.g., is 
more trained in logic. So being good for convincing cannot be solely a quality of 
the argument itself; it also depends on the addressee, on his epistemic state. The 
argument must be epistemically accessible to the addressee. 

2. Uncertain epistemological principles: Second, if we only used certain ways of 
cognizing, our knowledge would be utterly restricted; we could not make 
predictions or empirical generalizations, we could not explain on a theoretical 
level, or interpret texts or circumstantial evidence, or appraise our actions, etc. 
So, uncertain ways of cognizing and efficient but uncertain epistemological 
principles have been invented and are used in argumentation. This by itself 
already means giving up an ideal of alethic objectivity, because this kind of 
cognition no longer guarantees truth. In addition, such uncertain epistemological 
principles differ in probative force. Therefore, an uncertain argument may be 
sufficiently strong in one situation, e.g., when addressing laymen, but too weak 
in another situation, such as addressing an audience of experts with a scientific 
claim. So even this kind of sufficiency is not a feature of the argument itself, 
but of its use in a given situation. Finally, uncertain epistemological principles 
cannot guarantee the thesis's truth, and they are nonmonotonic: there may be 
different pieces of evidence (or an enlarged set of evidences) that lead to an 



226 Christoph Lumer 

incompatible thesis. Of course, epistemology provides means for deciding such 
conflicts. The relevant point of these solutions for our question is that uncertain 
cognition has to take into consideration all the available relevantj ustified beliefs. 
However, which justified beliefs are available does not depend on the argument, 
but, again, on the epistemic situation. 

3. Uncertain reasons: Third, once uncertain epistemological principals have got 
the epistemological placet, we also have to reckon with uncertain justified beliefs 
as the starting point of argumentation. We have to reckon with uncertain, in 
particular, probabilistic, premises or reasons, whose probabilities enter into the 
calculation of the thesis's probability. From an objective point of view, however, 
there are tendencies, propensities and relative frequencies, but no probabilities; 
propositions are either true or false even if we do not know what they are. On 
the other hand, for objectivity's sake to require of probabilistic arguments, e.g., 
that premises with a probability above 50% have to be true, and those with a 
probability of maximally 50% must be false, would distort the sense of such 
arguments. Such arguments are designed precisely for addressees who do not 
have a sufficient justification for such 100% beliefs; and the probability calculation 
of such arguments just starts with premise probabilities unequal to I or O. So an 
epistemological theory of argument should also allow good arguments-and not 
only their use-that are not objective in the respect that they contain probabilistic 
premises. 
These two conflicting requirements of epistemologically conceived 

argumentation have led to competing conceptions of good argumentation among 
champions of the epistemological approach: to objectivist conceptions (e.g., Biro 
1977; 1984; Siegel 1989; Siegel & Biro 2006) and to subjectivist conceptions 
(e.g., Feldman 1994; Sanford 1988; Sinnott-Armstrong 1999) of good 
argumentation. Similar divisions also exist among exponents of other approaches. 
The conflict is clearly evident in what the various theories require for premises-
ranging from truth to mere acceptance by the addressee---or in the requirements 
for the relation between the argument's reasons and its thesis-here the requirements 
range from inductive implication to, again, acceptance by the addressee. 9 

The solution to this tension, proposed by the Practical Theory of Argument, 
has already been intimated above. It consists in distinguishing (argumentative) 
validity and situational adequacy. Argumentative validity is a property of the argument 
itself; it guarantees that the requirements of objectivity are fulfilled; argumentative 
validity must imply that the thesis is objectively true or acceptable, which does not 
depend on a subject. Situational adequacy, on the other hand, is a relation between 
the argument and a situation (characterized mainly by the addressee's justified 
beliefs at the given time). Situational adequacy has to guarantee that the requirements 
of subjective epistemic accessibility are fulfilled: the argument's reasons must 
already be justifiedly believed by the addressee, he must know the epistemological 
principle, the complexity of inferential relations has to be adapted to his inferential 
abilities, etc. In rational convincing, the two sets of criteria have a "logical" order. 



The Epistemological Theory of Argument-How and Why? 227 

Only argumentatively valid arguments can be situationaIIy adequate; first of all the 
thesis's truth or acceptability has to be guaranteed, and then epistemic accessibility 
has to be secured. This has been explained instrumentally. We first have to look for 
instruments that in principle are suited to fulfil the required function, i.e., for 
which there is a situation in that they would fulfil the function; and among these 
we then have to choose the ones suited to the particular situation. Accordingly, the 
problem of criteria for good argumentation that try to get along with only one set 
of criteria is that they do not do justice to the two different functional roles of the 
two sorts of requirements: guarantee of acceptability and epistemic accessibility. 
Acceptability does not depend on subjects recognizing it, and epistemic accessibility 
is by no means guaranteed by acceptability. 

This is the principal idea. But the way argumentation functions and the interplay 
of the two sets of criteria have so far only been explained for deductive arguments. 
Deduction is a certain way of cognizing. Therefore, criteria for good deductive 
argumentation only have to deal with the first subjective challenge, i.e., epistemic 
accessibility of reasons and inferences (see above, DA4.2-3), but not with the 
other two: the problems caused by uncertain epistemological principles and uncertain 
reasons. If one wants to admit uncertain epistemological principles and arguments 
(and above I have argued that this is urgently required) one cannot uphold the 
original strong objectivity requirement for argumentative validity, namely that it 
has to guarantee the thesis's truth, that the premises have to be true and that they 
certainly imply the thesis. For uncertain types of arguments these requirements 
have to be replaced by weaker ones in such a way that argumentative validity 
guarantees only the thesis's acceptability, namely at least probability or verisimilitude. 
A very important feature of these weaker requirements is that, because of the 
nonmonotonicity of uncertain arguments, even argumentative validity not only 
depends on the acceptability of the specific reasons but also is relative to some 
presupposed database. The most important singular conditions for argumentative 
validity, then, are the following: I. The argument's underlying epistemological 
principle must be efficient, i.e., (mostly) leading to acceptable conclusions. 2. A 
given database may allow applying several epistemological principles or several 
inferences that lead to different conclusions, inconsistent with each other. (The 
database may, e.g., contain the information that Helen is female, 40 years old and 
a heavy smoker. And it may contain information about life expectancies of 40.-
year-old women and--different -life expectancies of 40-year-old female smokers.) 
Epistemology provides hierarchies of strengths of epistemological principles and 
rules for dealing with such situations. In case of conflicts, argumentative validity 
requires using the strongest epistemological principle or inference. (In the example, 
the inference with the stronger premises-'h is female, 40 years old and a heavy 
smoker; life expectancy for such people is 70 years'-is the stronger inference.) 
3. The set of premises taken from the database for the inference must be stable in 
the sense of including all (and only) the information from the database relevant to 
the thesis in question; e.g., adding further premises from the database would not 



228 Christoph Lumer 

lead to different conclusions. (We may, e.g., have testimony that Simon was the 
culprit yet have contrary results from genetic testing. In such a case, an inference 
to the best explanation has to include both these facts in the set of facts to be 
explained.) 4. (Prior) probabilities of the argument's premises must be correct 
with respect to the presupposed database. In addition, for rational convincing, 
situational adequacy requires that the argument's database be identical with that of 
the addressee, that the addressee be linguistically proficient, open-minded, attentive 
and discriminating, and more. 

Though argumentative validity of uncertain arguments is not objective in the 
sense of guaranteeing truth, it is not subjective in the sense of making argumentative 
validity relative to an addressee. (Far less is it rhetorical in the sense of requiring 
mere acceptance by the addressee.) One kind of relativity to the addressee here 
has been replaced by relativity to the database. This makes clear that argumentative 
validity does not depend on the addressee's subjective features, but only on the 
database, and that the reasons' probabilities do not depend on the addressee's 
arbitrariness but are justified by the respective database as well as by epistemological 
rules. The underlying idea is to approximate argumentative validity of uncertain 
arguments to objectivity in the strict sense as much as possible. 

Having explained the Practical Theory of Argument's proposal on how to 
conceive the relation between the objectivity and subjectivity of argumentation, 
this proposal can be defended in a more substantial way by comparing it with 
competing conceptions. An entirely objective account of epistemologically conceived 
good arguments would have to do withollt any reference to an addressee's epistemic 
situation. Such an account would perhaps work if the only way of cognizing were 
direct verification, which implies that for every proposition there is only one way 
of cognizing it, hence the same way for everybody; so there would not be much 
adjustment to the specific addressee. However, even in such a situation knowledge 
would differ intersubjectively, and the problem of epistemic accessibility would 
still have to be resolved: some people would know the premises to be true, others 
not. But above all, we have many more ways of cognizing than direct verification: 
indirect and uncertain cognition. And this makes some form of adjustment to the 
addressee's epistemic situation much more indispensable. Therefore, an entirely 
objective account of argumentation is a non-starter. And the existing objective 
accounts always contain some consideration of subjective requirements. lo To 
conceive argumentative validity as has been done here preserves as much as possible 
ofthe objectivist idea of guaranteeing truth or acceptability. But this kind of validity 
has to be supplemented by situational adequacy. 

If the particular epistemic situation in any case has to be taken into consideration, 
the subjectivist conception, just from the outset, is in a better position; and prima 
facie it seems to be obvious to define 'good argumentation' completely in subjective 
terms without any objective element. Feldman has followed this line, defining 
'good argument' this way: "An argument is a good argument for person S if and 



--
The Epistemological Theory of Argument-How.and Why? 229 

only if (i) S is justified in believing the conjunction of all the premises of the 
argument, (ii) S is justified in believing that the premises are 'properly connected' 
to the conclusion, and (iii) the argument is not defeated for S" (Feldman 1994, 
179)." Condition (iii) is a formula for resolving the problem of nonmonotonic 
reasoning that there may be justified inferences to a contradictory thesis that have 
precedence over the one in question. Feldman's definition is subjective in the sense 
that the goodness of an argument is defined relatively to a person. 

The prima facie advantage of a pure subjective account notwithstanding, I 
want to show that it has several disadvantages, which originate in the fact that it 
does not separate argumentative validity and the situational adequacy of arguments. 
\. In the case of an argument with an unrestricted (i.e., without probabilistic 
qualification) premise, p, that S at the time being was justified to believe in but that 
later turns out to be false, note that, according to Feldman's definition, one cannot 
say that the argument was and is bad or fallacious-though we usually do say so. 
We way the argument is fallacious though we did not recognize this; and it is 
fallacious because p is false. Our criticism is not that it is no longer justified to 
believe in p, but that p is false. So why not include truth of premise p in the 
conditions of a good argument? However, this per se is not a strong point, for 
perhaps we should just change our way of speaking. 2. Feldman's definition is 
short and elegant, but with the help of a trick: it strongly refers back to epistemology, 
which has to tell us (and S) when S is justified in believing in some premise and in 
"proper connections" between premises and a conclusion. 2.1.To begin with the 
last point: S is justified in believing in such a "proper connection" ifthere is such a 
connection and if S has recognized it. And then epistemology has to spell out what 
a "proper connection" is, e.g., that logical implication is a "proper connection". 
But this is exactly the same as what is spelled out in the respective condition of 
argumentative validity (see, e.g., DA2.2). So the (more or less) objective criteria 
have to be provided somewhere~ven if this is not made explicit in Feldman's 
definition. 2.2.S is justified in believing in probabilistic premises if(i) the premises' 
probabilities on the respective database are as specified in these premises-where 
epistemology, or more precisely: probability theory, has to tell how such probabilities 
have to be calculated-and (ii) if S has recognized this. Condition (i), though, is 
exactly what is said in the condition of argumentative validity of probabilistic 
arguments, whereas condition (ii) is part of the respective conditions for situational 
adequacy. So, again, the objective criteria show up. 2.3.There are too many 
possible ways to justifiedly believing in unrestricted (premises) to list them all. 
Objective criteria of argumentative validity here can only specify the target of 
these ways: the premises' truth. This again is what is said in the criteria of 
argumentative validity. But already this provides a criterion for criticism if the 
target has been missed-independently of the question whether the addressee was 
justified in believing the premise. 3. As is characteristic for a pure subjective 
account, Feldman provides only a unique criterion, which does not distinguish 



230 Christoph Lumer 

argumentative validity from situational adequacy. In the best case, his definition is 
extensionally equivalent to the conjunction of argumentative validity and situational 
adequacy. (Actually, Feldman's definition has a larger extension because it includes 
arguments with justified but false unrestricted premises (see above, point 1).12) 
But in any case, intensionally it does not distinguish two levels of argumentative 
goodness: the first guaranteeing acceptability (truth, high probability or 
verisimilitude), the second guaranteeing epistemic accessibility. These are quite 
different matters. Objective acceptability does not depend on cognizability and the 
respective situation. Only by distinguishing these things is it made clear why a 
good argumentation is good: it guarantees acceptability, and it does so in an 
epistemically accessible way. As a further aspect of the missing distinction, there 
is no distinction among quite different kinds offallacies. Is the argument good but 
it was only used in the wrong situation and could be adequately used in another 
situation? Or is the argument bad in itself? As a consequence, one may be seduced 
into some crude form of relativism, according to which knowledge is relative to 
subjects. Differentiating validity and adequacy knocks the bottom out of this 
relativism by proving that argumentative validity does not depend on the subject 
and by explaining the necessary subjective parts as epistemic accessibility. 4. But 
the biggest problem of Feldman's definition is that it does not tell us what the 
addressee has to check for to find out whether the thesis is true or acceptable or 
whether the argument shows what it is supposed to show. His criterion does not 
provide any assistance for assessing arguments and theses. And it cannot do so, 
because the objective part of the goodness of an argument is not worked out in the 
definition. 

The last point has to be examined a bit futher because it entails what makes up 
the central advantage of the proposal given in the Practical Theory of Argument, 
namely to develop a synthesis of the objective and the SUbjective account of the 
epistemologically-conceived goodness of argumentations. The central advantage 
is that the objective part of the criteria for argumentative goodness, i.e., the 
conditions for (argumentative) validity, specifies the conditions for the thesis's 
objective truth or acceptability; and by this it also specifies what the addressee has 
to check in order to cognize the thesis's truth or acceptability. This means this part 
explains why the argument is good; it provides an aid for cognizing the thesis's 
truth or acceptability; and it says what the subjective justification of one's belief 
ideally should consist in. The coincidence of these things also explains why checking 
an argument's (argumentative) validity with a positive result implies cognizing the 
thesis's acceptability (see above, sect. 4). An (argumentatively) valid argument is 
defined (by the respective validity conditions) in such a way that it consists of 
statements that the thesis's conditions of acceptability are fulfilled; if the addressee 
has positively checked that these statements are true he has cognized the thesis's 
acceptability. To make these points clearer by an example, take again condition 
DA2.2 for the argumentative validity of deductive arguments, which requires that 
the premises must logically imply the thesis. This is exactly (part of) what the 



The Epistemological Theory of Argument-How and Why? 231 

addressee has to check to cognize the thesis's truth. Compare this with Feldman's 
condition (ii): "5 is justified in believing that the premises are 'properly connected' 
to the conclusion". First, to cognize the thesis's acceptability it is not necessary 
for the addressee to check whether he is justified in believing, e.g., that the 
premises logically imply the thesis; he only has to check whether the premises 
imply the thesis. Second, to find out if he is justified in believing in the logical 
implication, the addressee has to check whether the premises imply the thesis. 
Third, because the target of asking oneself if one is justified in believing in the 
logical implication is different from that of asking oneself if the alleged implication 
holds, there might even be some difference in the respective outcomes. Even if 
with the last question one retrieves an already acquired justified belief (which in 
deductive argumentation is a rare case), this question invites more than the former 
to recheck whether the implication really holds. 

These advantages are those for the addressee. In addition there is an essential 
advantage for the arguer. The conditions of argumentative validity describe the 
argument's structure and thus are an aid, a manual, for constructing good arguments. 

The conditions of adequacy have quite a different role. They mostly have to 
guarantee the epistemic accessibility of the argument; they only specify the 
prerequisites for cognition, i.e., when the cognizing can be executed. If an argument 
is adequate for convincing an addressee, then he is in the position to carry out the 
cognition aimed at by the arguer, i.e., to check whether the reason-statements are 
fulfilled, etc.; if the argument is not adequate he simply cannot carry out the 
desired cognition. But to cognize the thesis's acceptability, the addressee does not 
have to examine whether or not the conditions of adequacy are fulfilled. So he has, 
e.g., to check whether the premises imply the thesis (OA2.2, part of argumentative 
validity), but there is no need to check whether this implication is sufficiently 
direct to be grasped by himself (OA4.3, part of situational adequacy). Either he 
grasps the implication, in which case he has already cognized this part of the 
thesis's acceptability conditions and no further work has to be done; or he does 
not grasp it, in which case he is not able to verify the objective implication, and 
stating that he does not grasp it does not help him to find out whether the implication 
holds. A mere subjective account of epistemologically conceived arguments blurs 
the distinction between these completely different requirements of good 
argumentation: between objective argumentative validity and subjective situational 
adequacy. 

7. Different Types of Arguments 

Deductive arguments are the only form of certain arguments. (They have been 
reconstructed from an epistemological point of view, e.g., by Feldman 1993, 61-
80; 94-100; Lumer 1990, 180-209.) As has already been stated, certain knowledge 
would be much too small a basis for orientating ourselves in the world. Uncertain 
forms of justified beliefs and of arguments are needed. These are based on 



232 Christoph Lumer 

epistemological principles other than the deductive, and they have to be differentiated 
according to the underlying epistemological principle. Genesis of justified belief 
arguments, e.g., are based on the genesis ofjustified belief principle: 'A proposition 
is true if it has been correctly verified'. The epistemological principle states in a 
general form the conditions under which a certain type of thesis is true or 
acceptable. In the appertaining arguments, the conditions of concretization of such 
a general principle for a specific thesis are stated to be fulfilled. For example: 'The 
traffic lights were still red when the car c started. This has been testified to by s 
who herself waited in her car beside c for green.' or: 'In the first century after 
Christ some of the Italic slaves used as simple farmhands had only 30 free days 
per year; see Columella, De re rustica 2, 11,8.' Genesis of justified beliefarguments 
also include arguments from testimony or from authority. Their essence is to 
describe or mention, in a more or less incomplete way, (i) how the thesis has been 
cognized by somebody and (ii) how this justified beliefhas been transmitted to the 
arguer. With respect to the last example, an expert, e.g., knows that Columella is a 
rather reliable and precise Roman author about agriculture, who himself had a 
farm and that his work has come down to us in some way ending up in carefully 
edited standard editions. So the addressee gets at least some pieces of information 
about the original cognition and about the transfer so that he can check them and 
see if they are correct, while he has to make probabilistic assumptions about the 
missing pieces. The multitude of such assumptions makes genesis of justified 
belief arguments a rather weak sort of arguments. 

Genesis of justified belief arguments are only a first type of probabilistic 
argument. Their argumentative form (including criteria of argumentative validity 
and/or situational adequacy) and epistemological force have been studied (by 
Feldman 1993, 216-232; 418; Lumer 1990, 246-260). Another special form of 
probabilistic arguments that has been reconstructed from an epistemological point 
of view are interpretative arguments, which try to establish the causes for known 
facts and circumstantial evidence by an inference to the best explanation based on 
Bayes's Theorem (see Lumer 1990, 221-246; Lumer 1992). In these and other 
cases of probabilistic arguments (see, e.g., Korb 2003) the underlying 
epistemological principles are those of probability theory. Some statistical types of 
arguments, whose epistemological principles are obviously those of statistics, as 
well as causal arguments, have been reconstructed by Feldman (1993, 232-327). 
Practical arguments about the desirability of states of affairs and about optimum 
courses of action, which list the object's pros and cons and try to assess and 
weigh them, have also been reconstructed from an epistemological point of view 
(by Feldman 1993, 351-354; 420; Lumer 1990, 319-433). The latter arguments 
are based, among others, on utility theory and the theory of practical rationality as 
their epistemological principles. Pascal arguments, which are based on rational 
decision theory and are arguments under complete uncertainty in favour oftreating 
a thesis as true, have also been reconstructed epistemologically (Lumer 1997). 
Deductive, probabilistic and practical arguments are basic forms of arguments 



The Epistemological Theory of Argument-How and Why? 233 

and not reducible to each other; interpretative and genesis of justified belief 
arguments are special types of probabilistic arguments, whereas Pascal arguments 
are special types of practical arguments. 

This list is not intended to be exhaustive. The important points for the present 
discussion are these. First, all argumentatively valid arguments are based on and 
appeal to some efficient epistemological principal, which is justified in epistemology 
and which guarantees the thesis's truth or acceptability by its relation to the thesis's 
truth conditions. Second, different types of argument can be distinguished according 
to their underlying epistemological principle. Third, the epistemological principles 
used in non-deductive arguments are more complicated than the deductive principle 
and thus lead to more complex forms of argument, which cannot be reduced to 
deductive arguments. A first structural enlargement is that criteria for the 
argumentative validity of all uncertain forms of argument include a reference to 
some database, which, according to the adequacy conditions, then has to be that 
of the addressee. This means the argument's validity is relative to this database and 
does not only depend on the argument's premises. This feature has already been 
dealt with in the last section. A further structural enlargement is specific to practical 
arguments. Practical arguments, like other arguments, are to guide the addressee 
to cognize the (epistemic) acceptability of the thesis. This epistemic part of practical 
arguments already leads to new complications, because the object of the value 
judgement may have huge numbers of possibly relevant consequences and 
implications, which even in principle we cannot all justifiedly believe and survey. 
However, the practical part of practical arguments is a still more fundamental 
enlargement. What does this practical part consist in? It has to do with practice: 
finally believing in the thesis (i.e., a value judgement) has to motivate the addressee 
to some degree with respect to the value object, to materialize the value object, to 
protect it, etc., or to destroy, to flee from it, to abandon it, etc. The addressee may 
believe as strongly and justifiedly as one wants in the thesis, if he does not have 
such a kind of motivation, the argument has failed as a practical argument. If 
practical arguments can guide cognition only like other arguments (namely to 
guide checking whether the thesis's acceptability conditions are fulfilled) then the 
practical part of practical arguments must be a question of thesis selection. There 
may be countless true theses about the object or action in question, all well-defined 
and many of them cognizable so that an argumentatively valid and adequate argument 
for them exists. The important feature that makes one or only a small number of 
these arguments into practical arguments is that their theses have such a content 
that believing in them generates the respective motivation. Arguments and reasons 
fulfilling this motivational requirement in practical philosophy have been called 
"internal" (see, e.g., Williams 1979). However, motivation is an empirical, causal 
effect. This implies that being a good practical argument depends on the addressee's 
or, more generally, on human, motivational make-up; and criteria for the 
argumentative validity and adequacy of practical arguments have to be based on 
the respective psychological findings. This point cannot be worked out further in 



234 Christoph Lumer 

this article. What is important here is that constructing validity and adequacy criteria 
for various types of arguments does not only require basing them on epistemology 
but eventually, when it comes to practical arguments, transcends even this base 
and additionally requires foundation in theories of practical rationality and in the 
psychologies of motivation and decision making. 

8. General Criteria for Argumentatively Valid and Adequate 
Arguments 

With the elucidations given so far, the general definition of '(argumentatively) valid 
argument' and with it the general criteria for good arguments (first in a rough form 
then in an extended and precise form) should be understandable. 

An argument x is (argumentatively) valid iff (i) there is a (general) 
epistemological principle e and a criterion C that is a concretization of e for the 
thesis q of x (so c says: q is acceptable if conditions C I' ... , C m are fulfilled), and the 
reasons of the argument are identical with the conditions c

1
' ... , c

m 
of C or with a 

part of them; (ii) e is efficient (see above sect. 5); (iii) the conditions c
1

' ... , c
m 

of 
C are fulfilled; (iv) there is at least one person who does not yet justifiedly believe 
the thesis and who if confronted with x would be guided by x to cognize the 
acceptability of the argument's thesis. 

The central point of this definition is that the reasons of a valid argument 
truthfully state some propositions to be true that are identical to all or part of the 
concretized conditions of an efficient epistemological principle, concretized for 
the thesis. One of the infinite concretizations of the deductive epistemological 
principle e.g. is: "'Socrates is mortal" is true if 1.1. both "all humans are mortal" 
and 1.2. "Socrates is human" are true and if 2. these two propositions imply 
"Socrates is mortal".' Everyday arguments would omit two of the three possible 
reasons thus being condense, e.g., to: 'Socrates is mortal since he is human.' 

An argument in a broad sense is a valid argument or an argument believed or 
stated by someone to be valid. 

A valid argument is adequate for rationally convincing a listener (hearer) h if h 
is not yet sufficiently convinced of the argument's thesis, if h knows at least 
implicitly the underlying epistemological principle, if h justifiedly believes (which 
includes that in this very moment h cognizes) the conditions of the concretization 
C to be true, if h is able to add the most important missing conditions of acceptability, 
if the argument is sufficiently strong for h's purposes, and if h has no relevant 
knowledge about the thesis above the database presumed in the argument. 

For the various types of arguments more specific definitions should be developed 
that rely on the respective epistemological principles; such definitions are much 
more helpful for argumentative practice. Developing them here, however, would 
exceed the limits of this paper. J3 The rest of this section provides more exact 
versions of the general conditions just introduced. 



The Epistemological Theory of Argument-How and Why? 235 

x is a valid argument (i.e., an argument in the narrow sense) = 

AO: Domain of definition: x is a triple <ro,i,q>, consisting of 

AO.I: a set rO of judgments/statements r
l
, r

2
, ••• , r

n
, 

AO.2: an indicator i of argument, and 

AO.3: a judgment q; 

r
l
, ... , r

ll 
(the elements of rO) are called the "reasons for q" and q is called "the 

thesis of x". 

Ai: Indicator of argument: i indicates that x is an argument, that r
l
, r

2
, ... , r

ll 
are 

the reasons and that q is the thesis of x; in addition i can indicate the type of 
argument, i.e., the epistemological principle on which the argument is based. 

A2: Guarantee of acceptability: There is an epistemological principle e and a criterion 
c for the acceptability fulfilling the following conditions: 

A2.I: Efficient (epistemological) principle: the epistemological principle e is 
efficient; and 

A2.2: Concretization (of the principle): the criterion c is a concretization of the 
principle e for the thesis q (so c says: 'q is acceptable if c

l
' ... , cm'), and the 

reasons r
l
, r

2
, ... , rn are judgments claiming of at least a part of the conditions (c

l
, 

c
2
, ... , cm' with m~n) of c that they are fulfilled (these reasons may have a probabilistic 

form and refer to a database like this: r
i 
= 'the probability of b

i 
on the database dis 

p/); and 
A2.3: True reasons: all conditions (c

l
' c

2
' ... , cm) of c are fulfilled. 

A3: Adequacy in principle: x fulfils the standard function of arguments; i.e., there 
is a subject sand a.time t for which holds: 

A3.I: the subject s at the time t is linguistically competent, open-minded, 
discriminating and does not know a sufficiently strong justification for the thesis 
q;and 

A3.2: if at t x is presented to sand s closely follows this presentation this will make 
s justifiedly believe that the thesis q is acceptable; this process of cognition will 
work as follows: s, using e and c, will recheck-among others-those conditions 
(among c

l
' c

2
' ... , cm) for the acceptability of the thesis q which are claimed to be 

fulfilled in r
l
, r

2
, ... , r

n
, thereby coming to a positive result. 

x is an argument (in the broad sense, which includes invalid arguments) = 

A4.0: Domain of definition: The domain of definition is the same as that of valid 
arguments. 

A4.I: Valid argument: x is a valid argumentation, or 

A4.2: Seemingly valid argument: there is a person s and a moment t with s at t 
believing or (explicitly or implicitly) holding the view that x is a valid argument. 

A valid argument x is adequate for rationally convincing an addressee h 
(hearer) at t of the thesis (q) of x iff condition A5 holds: 



236 Christoph Lumer 

A5: Situational adequacy: 

A5.1: Rationality of the addressee: The addressee h (at t) is linguistically competent, 
open-minded, discriminating and does not have a sufficiently strong justification 
for the thesis q. And 
A5.2: Argumentative knowledge (of the addressee): A5.2.1: The addressee h at t 
kDows at least implicitly the underlying epistemological principle e ofthe argument 
x; and A5.2.2: at t he (h) is able to develop the criterion c of acceptability (which is 
intimated in x) by means of his knowledge of the principle e if all the reasons of an 
ideal version of x are presented to him. And 

A5.3: Acceptance of the reasons: The addressee h at t justifiedly believes that the 
propositions c

l
' c

2
' ••• , c

m 
are true, with c

l
' c

2
' ••• , c

m 
being the conjuncts of the 

antecedent of the criterion c of acceptability (intimated in x). And 

A5.4. Explicitness.' If in the reasons of x not all conditions (c
l
, c

Z
' ••• , cJ of the 

criterion c of acceptability (intimated in x) are claimed to be fulfilled the addressee 
hat t is able to add the most important conditions of acceptability. And 

A5.5. Sufficient argumentative power: A5.5.1: The criterion c of acceptability 
(intimated in x) together with the subjective probabilities of the addressee (h at t) 
that the conditions of c are fulfilled provide a sufficiently high degree of probability 
of the thesis (q of x}-sufficiently high according to the desires of the addressee 
(h at t); and A5.5.2: in case of a nonmonotonic argument the database of the 
addressee h at t does not contain relevant justified beliefs which are not enclosed in 
the database presumed by the argument. 

9. Justification of the Epistemological Approach to Argumentation 

The epistemological approach to argumentation, in the specific form of the Practical 
Theory of Argument, so far has been sufficiently expounded as now to be justified. 
The epistemological approach here has been characterized by a certain conception 
of the standard function of arguments: to lead the argument's addressee to justified 
belief. The most important competing determinations are: 1. rhetorical: the aim or 
function of argumentation is persuasion, i.e., to cause or to increase acceptance of 
the thesis (Perelman & Olbrechts-Tyteca 1958, 5; Hamblin 1970, 241; see also 
Blair & Johnson 1987, 48); 2. consensualistic: the aim or function of argumentation 
is consensus, i.e., to produce a shared belief (Habermas 1981 I, 48) or-in van 
Eemeren and Grootendorst's terms-to "solve conflicts of opinion" (van Eemeren 
& Grootendorst 1984, 1 et passim; van Eemeren et al. 1996,275; van Eemeren & 
Grootendorst 2004,5; 56). Consensualistic approaches mainly take discussions as 
their principal object of study (and van Eemeren and Grootendorst use 
"argumentation" mostly in this sense). If one concentrates on the argumentative 
aspect of such discussions the main difference between the consensualistic and 
the rhetorical conceptions of argumentation is that, according to the former, in 
good argumentation the arguer must believe what she says, i.e., she must accept 



The Epistemological Theory of Argument-How.and Why? 237 

the whole argument for herself, which is not required in rhetorical theories. 

The epistemological approach to argumentation now is justified by its various 
advantages: 

1. More truths: It designs good arguments as instruments that guide people to 
acquire justified belief.lfsuch arguments are used in the proper way, the output 
is not only that the addressee believes something new but that he justifiedly 
believes it. This implies that this belief is acceptable: true or at least truth-like or 
probably true. This is guaranteed by relying on epistemological principles that 
are justified by their particular relation to the truth definitions of the respective 
theses. And true (and often also truth-like) beliefs are helpful for orientating 
ourselves in the world and for choosing the best course of action; they are 
much more helpful than false beliefs. We can reach our ends because we 
justifiedly believe what the consequences of our actions are; and we can choose 
the best end and option because we justifiedly believe their implications and to 
what degree these fulfil our criteria for good ends and actions. In addition, 
admitting uncertain arguments extends the set of knowable propositions to a 
degree that is necessary and sufficient for practical matters. It does so with the 
help of risky procedures, which, however, are still efficient. Of course, as a 
consequence, even the output of epistemologically good argumentation may be 
a false belief. But, because of being based on epistemological principles, the rate 
of true or truth-like beliefs among all practically important beliefs resulting from 
epistemologically designed argumentation is the highest obtainable for humans. 
In any case, it is much higher than the rates for argumentation designed according 
to the rhetorical or consensualistic approach, because these approaches lack 
any direct connection to truth conditions. '4 

2. Cumulation of knowledge: Beliefs acquired by epistemologically designed 
argumentation are supplemented by their respective justification. This makes 
the sources of the beliefs retrievable and implies information about the strength 
of justification. As a consequence, in cases of conflicting beliefs, the one with 
the weaker justification can be given up and with it further beliefs depending on 
the corrected belief, etc. All this amounts to a process of replacing false beliefs 
with true ones or at least statistically increasing the rate of true beliefs. This 
holds because uncertain epistemological principles have to integrate the person's 
complete database into one coherent picture, which, because of the growth of 
this database, excludes more and more simplistic interpretations that were 
compatible only with the smaller database. This means that the epistemologically 
organized process of belief acquisition is cumulative. Nothing of this holds for 
rhetorically or consensualistically designed argumentation, because neither is 
bound to a cumulative system of knowledge change. Even if it allows belief 
changes, these changes are not related to truth; so, occasionally, false beliefs 
will be replaced by true ones, but much more frequently false or true beliefs will 
be replaced by false ones. 



238 Christoph Lumer 

3, Usability as subjective justification: Epistemologically conceived arguments 
represent the ideal form of propositional subjective justifications; so subjects 
may directly store the argument as the justificatory part of their justified belief. 

4. Transfer of justified belieft, Epistemologically designed argumentation of an 
arguer addressed to another person transfers justified belief as such. 

5, Perspective of a truth seeking person.' The epistemological approach designs 
arguments from'the perspective of an addressee who wants to examine the 
thesis's acceptability. He checks the reasons to ascertain whether they are 
sufficiently strong for rationally accepting the thesis. Epistemological 
argumentation theory provides criteria for such examination. (See Feldman 1994, 
175; Lumer 1991,98; Meiland 1989,194.) It thus supports the rational and 
emancipated subject. Rhetorical approaches, on the other hand, tend to 
instrumentalize the addressee by inducing beliefs that do not have the advantage 
of being bound to truth and that are chosen by the arguer for his persona! 
reasons. And consensus theories tie each subject to other people's opinion, 
without offering any epistemic advantage. 5.1. A side effect of taking the 
addressee's perspective is that the respective arguments are also suited for 
mono logic use, for finding out the truth, inventing hypothetical arguments, etc. 

6, Precise criteriafor assessing arguments: The epistemological approach provides 
(or at least aims at providing) precise criteria for assessing arguments according 
to their epistemological value. These criteria can be used to design arguments 
as well as to decide on them. Consensus theories, on the other hand, do not 
have anything to offer in this respect; they stop after having provided procedural 
rules. And rhetorical approaches either provide only classifications of arguments 
without precise prescriptions, or they propose recipes that instrumentalize the 
addressee. 
This justification is not linguistic, but practical, or more precisely, instrumentalist 

-and this is the reason for the name "Practical Theory of Argument". It does not 
appeal to the common or "true" sense of "argument" or "argumentation" but lists 
instrumental advantages of the objects defined before. The justification does not 
rule out that we also may develop, e.g., a rhetorical theory of persuasive speech 
and that in certain situations it would be instrumentally better and morally legitimate 
to use instruments designed by such a theory, i.e., well-designed rhetorically 
persuasive speech-think, e.g., of someone who tries to convince some terrorist 
to surrender but who thinks that the most effective premises he could use, which 
come from the terrorist's confused beliefs, are all false. ls However, although these 
rhetorical instruments will have other advantages, they will not have those listed 
above. 

This liberal view about these different types of instruments and approaches 
notwithstanding, some priorities should be clear. First, epistemologically conceived 
arguments do and should prevail in most domains of our lives: of course in science, 
but also in political decisions, courts of law and daily life, because here we are 
interested in truth and in finding out best solutions. And such arguments provide 



The Epistemological Theory of Argument-How and Why? 239 

this much better than, e.g., mere rhetoric (see advantages I and 2). Second, even 
if someone simply intends to make an addressee believe some proposition, of 
course epistemologically-conceived arguments could do this work too. Sometimes, 
though, mere rhetoric could be the better instrument for achieving this goal. But 
even in this case, rhetorical success often is parasitic upon valid argumentation, 
because the addressee has the vague impression ofhaving been rationally convinced. 
Third, historically the terms "argument" and "argumentation" have been used for 
the kinds of speech and meaning that are bound to truth. An epistemological approach 
to argumentation elaborates this more or less vague understanding. So the concepts 
finally defined by this approach should be the concepts of 'argument' and 
'argumentation' -and not, e.g., the objects defined in rhetorical or consensualistic 
approaches. Of course, one could invert the meanings of these terms without 
much detriment because this is a secondary question that does not change the 
concepts and the value of the objects; but it would be a break with the linguistic 
tradition. 

Acknowledgements: I would like to thank Tony Blair, Ralph Johnson and Christopher 
Tindale for their very valuable comments on this paper. 

Notes 

I This theory was first elaborated in Lumer J 990. An English exposition of some of its main 
features is Lumer 1991. Some general extensions of this theory are provided in turner 2000a (a 
general theory offallacies), Lumer 1988 (embedding of monologic argumentations in dialogical 
disputation for collectively enquiring into truth), Lumer 1995 (methodology and type of theory). 
2000b (relation to logic). Lumer 2003 (interpreting argumcnts with the aim of evaluating thcm 
critically). The present paper advances some parts of this theory in ordcr to ans\\(:r important 
questions in the current debate about the epistemological approach. 
1 Propositions in the narrow sense roughly are the meanings ofthat-c\auses Arguments. however. 
are not expressed by that-clauses but by assertive sentences. whose meaning, in addition to the 
proposition, contains the assertive mood. Unfortunately, both English expressions for this meaning. 
are ambiguous: "statement", in addition to these intensional objects. refers to assertive sentences 
and to assertive acts; "judgement", in addition to the intensional object, refers to the mental act of 
judging. 
J To "cognize'" here shall mean the process by which justified belief is acquired: probably it is the 
expression that most unambiguously refers to th is process. 
4 Here, rationally convincing has been adopted as the standard function of arguments; and this is 
reflected in the terminology used in the following; it speaks ofthe "arguer'" and the ·';Jddressec". 
etc. But there is no problem, and nothing substantial will be changed. if autonom"llS cognition 
were be taken as the standard function of arguments. From the standpoint of rationally convincing. 
autonomous cognition with the help of arguments may be regarded as a limiting case. where arguer 
and addressee are identical: the monologue has become a soliloquy. This is possible because the 
standards for rationally convincing to be proposed below do not require that the arguer already 
justifiedly believe in the thesis. From the point of view of autonomous cognition. an argument 
presented by another person is only an hypothesis for a good argument (other people arc heuristic 



240 Christoph Lumer 

devices so to speak), which then has to be checked for its validity. The kernel of both autonomous 
cognition and rationally convincing is that the subject and the addressee, respectively, check the 
conditions for the thesis's acceptability; so the kernel in both cases is convincing oneself(see the 
following section). 
5 The complete criteria for deductive arguments are given in Lumer 1990, 187-189. These complete 
criteria, apart from being more precise, in particular also deal with simplified and abridged arguments 
like enthymemes. 
(, In order not to complicate the principal explanation, it is here stipulated that the addressee 
knows the premises to be true. Of course, deductive arguments can also start from merely 
acceptable premises. In that case, the argument is no longer certain and monotonic, so that 
precautions have to be taken to exclude that other information of the addressee defeats the thesis 
(see below). 
7 The more a person is trained in this respect. the better she can distinguish various types of 
arguments, and the more precisely she knows what conditions exactly have to be fulfilled. To have 
an explicit knowledge of such principles and their argumentative use is one ofthe key qualifications 
for good argumentation. 
'The lack of clarity about the epistemological principle used is one of the big problems of present 
argumentative practice in general. To improve this practice, at least in difficult cases, one should 
even include the respective epistemological principle explicitly in the argument's reasons. This 
means the remaining arguments would have these principles as implicit reasons. A different 
solution, which presupposes more explicit knowledge about epistemological principles, is to 
invent and use more type-specific indicators of argument. 
9 A nice short overview of some diverging sets of criteria resulting from attempts to deal with the 
just-discussed tension is given in Govier 1992, 393f 
III In their theory of begging the question, Siegel and Biro, e.g., distinguish between a good 
argument and its use in a given situation, where this use may be pointless (Siegel & Biro 2006, end 
of sect. 2). This is rather close to the distinction made here, i.e., the distinction between 
argumentative validity and situational adequacy. 
II Feldman explains that "justified in believing x" does not imply that the person believes x but 
only thatx is justified (Feldman 1994, 177). I think this specification goes in the wrong direction 
because without belief in the premises, epistemic accessibility of the premises would no longer be 
guaranteed. There might be a very big step from beingjustified in believing, to justifiedly believing, 
and often a step too big for the: addressee. Belief in the inference, on the other hand. in adequate 
argumentation usually is acquired only in the moment of argumentation itself But even here, 
problems may arise for making the step from being justified in believing in the inference to 
justifiedly believing it. But this is a minor point that does not affect the main difference between 
an objective and a subjective account. So in the following I will ignore these problems. 
12 One could try to make an extensional distinction between argumentative validity and adequacy 
in Feldman's terms in a very simple manner. One could define '(argumentatively) valid argument' 
like this: x is an argumentatively valid argument if there is a person s for whom x is a good 
argument. The extension of this term is again much larger than the extension of what here has been 
defined as '(argumentatively) valid argument'. It would, e.g., include a vast quantity of deductive 
arguments with false premises. But this extensional approach does not resolve the intensional 
problem. 
13 1 have defined some specific types of argument elsewhere: deductive arguments (Lumer 1990, 
187-189), interpretative arguments (to the best explanation) (Lumer 1990, 237-244), genesis of 
justified belief arguments (Lumer 1990, 258-260), practical arguments for value judgments (Lumer 
1990,362-366), Pascal arguments for theoretical theses under uncertainty (Lumer 1997, 339). 
14 Some consensualistic or rhetorical approaches (Perelman, Habermas, Tindale (1999, e.g., 117-
120» take 'convincing a universal audience' or 'universal consensus' as criterion for truth or for 



The Epistemological Theory of Argument-How and Why? 241 

particular argumentative quality. 1. The main problem with this proposal is that it lacks connection 
to substantial definitions and criteria of truth. Why should consensus have anything to do with 
how the world is? 2. Actual universal consensus admittedly never exists. If it existed, consensus 
theories of truth could not explain it. But the most obvious explanation would be that this 
consensus originates from cognizing the truth, where 'truth' then must be defined independently 
of consensus. 3. If universal consent never exists, appeal to it can only be de jure (Perelman & 
Olbrechts-Tyteca 1958.41), i.e., that the respective proposition should find universal consensus. 
But in order to establish what should find universal consent, one again needs independent criteria. 
And, of course, these should be alethic and epistemological. (For a more extensive critique of 
particularly Habermas's proposal, see Lumer 1990,291-296.) 
III see much less room for a consensualistic theory of"argumentation" that is not epistemologically 
conceived. (An epistemologically conceived consensual ism could, e.g., strive for an argumentatively 
justified consensus as a means of integrating socially distributed knowledge and of minimizing 
errors of cognition, thus increasing one's rational certainty about a proposition. I have developed 
such an approach in Lumer 1988.) In non-epistemological consensualistic discussions, each 
participant, as a rational person,for herseifshould strive for justified belief, and thus should be 
argumentative. On the other hand, she should not do so with respect to the other participants-
otherwise we would have a completely epistemologically-conceived discussion. I do not see a 
coherent function for such discussions. Either they collapse into a rhetorical conception by giving 
up the consensus requirement and allowing strategic behaviour. Or they collapse into an 
epistemological conception by giving up the substantial liberality of arguments and requiring that 
only epistemologically designed arguments may be advanced in the discussion. 

References 

Biro, John I. (1977). Rescuing 'Begging the Question'. Metaphilosophy 8: 257-271. 

-. (1984). Knowability, Believability, and Begging the Question. A Reply to Sanford. 
Metaphilosophy 1'5: 239-247. 

Biro, John I., Harvey Siegel (2006): In Defense of the Objective Epistemic Approach to 
Argumentation. Forthcoming in: Informal Logic, 26. 

Blair, 1. Anthony. (2004). Argument and Its Uses. Informal Logic 24: 137-151. 

Blair, J. Anthony, Ralph H. Johnson. (1987). Argumentation as Dialectical. Argumentation 
1:41-56. 

Eemeren, Frans H. van, Rob Grootendorst. (1984). Speech Acts in Argumentative 
Discussions. A Theoretical Model for the Analysis of Discussions Directed towards 
Solving Conflicts of Opinion. Dordrecht; Cinnaminson: Foris. 

-. (2004). A Systematic Theory of Argumentation. The pragma-dialectical approach 
Cambridge: Cambridge University Press. 

Eemeren, Frans H. van, Rob Grootendorst, Francisca Snoeck Henkemans, [et af.]. (1996). 
Fundamentals of Argumentation Theory. A Handbook of Historical Backgrounds 
and Contemporary Developments. Mahwah, NJ: Lawrence Erlbaum. 

Feldman, Richard. (1993). Reason and Argument. 2nd Edition. Upper Saddle River, NJ: 
Prentice-Hall 1993; 1999. 

-. (1994). Good Arguments. In: Frederick F. Schmitt (ed.): SocialiZing Epistemology. 
The Social Dimensions of Knowledge. Lanham, MD: Rowman & Littlefield. Pp. 159-
188. 



242 Christoph Lumer 

Govier, Trudy. (\ 992). What is a Good Argument. Metaphilosophy 23: 393-409. 

Habermas, JUrgen. (1981). Theorie des kommunikativen Handelns. Band 1: 
Handlungs-rationalitat und gesellschaftliche Rationalisierung. Frankfurt: 
Suhr-kamp. 

Hamblin, C[harles] L[eonhard]. (1970). Fallacies. London: Methuen. 

Johnson, Ralph H. (2000). Manifest Rationality. A Pragmatic Theory of Argument. 
Mahwah, NJ/London: Lawrence Erlbaum. 

Korb, Kevin B. (2003). Bayesian Informal Logic and Fallacy. Informal Logic 23: 41-70. 

Lumer, Christoph. (1988). The Disputation. A Special Type of Cooperative Argumentative 
Dialogue. Argumentation 2: 441-464. 

-. (1990): Praktische Argumentationstheorie. Theoretische Grundlagen, praktische 
Begriindung und Regeln lVichtiger Argumentationsarten. Braunschweig: Vieweg. 

-. (\ 99 I): Structure and Function of Argumentations. An Epistemological Approach to 
Determining Criteria for the Validity and Adequacy of Argumentations. In: Frans H. 
van Eemeren ret al] (eds.): Proceedings of the Second International Conference on 
Argumentation. Orga-nized by the International Society for Ihe Study of 
Argumentation (ISSA) at the Univer-sily of Amsterdam, June 19-22, 1990. Pp. 98-
107. Amsterdam: SicSat. 

-. (1992). Handlungstheoretisch erklarende Interpretationen als Mittel der semantischen 
Bedeutungsanalyse. In: Lutz Danneberg; Friedrich Vollhardt (eds.): Vom Umgang 
mil Lileralur und Lileraturgeschichte. Positionen und Perspektiven nach der 
"Theoriedebatte ". Pp. 75-113. Stuttgart: Metzler. 

-. (1995). Der theoretische Ansatz der Praktischen Argumentationstheorie. In: Harald 
Wohlrapp (ed.): Wege der Argumentationsforschung. Pp. 81-101. Stuttgart-Bad 
Cannstatt: frommann-holzboog. 

-. (1997). Practical Arguments for Theoretical Theses. Argumentation 11: 329-340. 

-. (2000a). Reductionism in Fallacy Theory. Argumentation 14: 405-423. 

-. (2000b). Argumentationstheorie und Logik. In: Geert-Lueke Lueken (ed.): Formen 
der Argumentation. Pp. 53-71. Leipzig: Leipziger Universitatsverlag. 

-. (2003). Interpreting Arguments. In: Frans H. van Eemeren; J. Anthony Blair; Charles 
A. Willard; A. Francisca Snoeck Henkemans (eds.). Proceedings of the Fifth 
International Conference of the International Society for the Study of 
Argumentation. Pp. 715-719. Amsterdam: SicSat. 

Meiland, Jack W. (1989). Argument as Inquiry and Argument as Persuasion. 
Argumentation 3: 185-196. 

Perelman, Ch[aim], L[ucie]. Olbrechts-Tyteca (1958). La nouvelle rhlitorique. Traite de 
I 'argumentation. 2 vol. Paris: Presses Universitaires de France. 

Sanford, David H. (1988). Begging the Question as Involving Actual Belief and 
Inconceivable Without It. Metaphilosophy 19: 32-37. 

Siegel, Harvey. (1987). Relativism Refuted. Dordrecht: Reidel. 

-. (1989). Epistemology, Critical Thinking, and Critical Thinking Pedagogy. 
Argumentation 3: 127-140. 

Sinnott-Armstrong, Walter. (1999). Begging the question. Australasian Journal of 
Philosophy 77: 174-191. 



The Epistemological Theory of Argument-How and Why? 243 

Tindale, Christopher W. (1999). Acts of Arguing. A Rhetorical Model of Argument. 
Albany, NY: State University of New York Press. 

Williams, Bernard. (1979). Internal and External Reasons. In: Ross Harrsion (ed.): Rational 
Action. Studies in philosophy and social science. Pp. 17-28. Cambridge; London: 
Cambridge U.P. Reprinted, e.g., in: Elijah Millgram (ed.): Varieties of Practical 
Reasoning. Pp. 77-97. Cambridge, MA.; London: MIT Press 2001. 

Christoph Lumer 
University of Siena 

Department of Philosophy 
Via Roma. 47 
53100 Siena 

1taly 

lumer@unisi.it